Artificial Intelligence (AI) is no longer a buzzword; it’s reshaping industries and economies globally. From optimizing supply chains to transforming healthcare and transportation, AI holds vast potential, but how will this affect greenhouse gas emissions?

While direct impacts on emissions are currently small, the rapid expansion of AI could see energy consumption skyrocket, according to researchers in a new commentary published in Nature. 

Owen Gaffney, co-founder of the Exponential Roadmap Initiative, and colleagues Amy Luers, Jonathan Koomey, Eric Masanet, Felix Creutzig, Juan Lavista Ferres, and Eric Horvitz, say it’s imperative that decisions around AI development and use are made with sustainability in mind.

Calling on the need to explore scenarios on AI impact on greenhouse gas emissions, the researchers suggest a set of policy-driven scenarios to quantify the effects that AI expansion could have on the climate under a range of assumptions.

“We need to bring climate researchers, policymakers and the large tech companies together in an international consortium specifically exploring scenarios about AI impact on the planet,” says Gaffney.

A set of scenarios would be used to better understand how AI expansion might affect emissions, and ensure AI’s evolution aligns with the need to radically reduce emissions and scale up climate solutions to ensure a livable planet for all.